Immobilized enzyme electrodes

Preferred enzyme electrodes are glucose oxidase electrodes containing glucose oxidase adsorbed or immobilized on a substrate surface.

The transfer of the enzyme electrode electrode to the amperometric weight of the catalytic activity of the enzyme and to the nerve is then immobilized on the substrate and the enzyme, adsorbed onto the electrically charged portion of the electrode, which is preferably used to reduce the risk. Whereas, in the case of such products or adsorbents, the same conditions apply to the binding of the skins, In this case, the platinum group metal, for example, is used as a porous substrate to adsorb such enzyme adsorbers or immobilizers and reactants, and in an interesting heterogeneous form with a resin-free coil or a graphite particle, the platinum group metal particles being dispersed. In the case of secondary substrates, a platinum adsorbent resin based on a colloidal platinum adsorber can be used as a particle and a bundle account for the substrate under the bonding of the synthetic resin, which is polystyrene. However, the enzyme electrode electrode of the glucosoxide electrode is impregnated with glucosoxide adsorbed or immobilized on the substrate.

8851 5

Enzyme Electrode - Enzyme Electrode

This invention relates to enzyme electrodes comprising an enzyme immobilized on an electrically conductive substrate and which react amperometrically to the catalytic activity of the enzyme in the presence of its corresponding substrate. In particular, but not exclusively, the invention relates to enzyme electrodes that can be used to determine glucose levels both in vitro and in vivo, comprising an electrically conductive substrate immobilized with an oxidoreductase, e.g., glucose oxidase, which electrode reacts amperometrically to the catalytic activity of the immobilized enzyme when glucose-containing sample.

The advantages of amperometric biosensors containing the enzyme as a biocatalyst are summarized in relatively detail by Aston and Turner, (1984) Biotech. Genet. Eng. Rev. (ed. G. Russell), 1, 89-120, Intercept, Newcastle-upon-Tyne, and Davis, G., (1985) Biosensors, 1, 161-178. They vary in the mode of signal transmission and the different types can be roughly classified as (a) those in which the electrical response is due to oxidation of the enzyme reaction product by the electrode; (b) mediator aids, in which electrons are transferred from the enzyme to the electrode by means of an redox agent, or (c) direct electron transfer (DET), where no such intermediary aid is required).

Group (s)

This type can be illustrated by the action of certain oxidases (eg glucose oxidase, alcohol oxidase) which produce hydrogen peroxide according to the following reaction:

8851 5 substrate + O2 --- [oxidase] ---> oxidized product + H 2 O 2

In this method, the peroxide is oxidized by an electrode with a certain potential:

B₂O₂ -------> 0₂ + 2H⁺ + 2e

The electrical signal is formed by the transfer of electrons from the peroxide to the electrode and under favorable conditions the enzyme-catalyzed current is directly proportional to the analyte concentration.

Numerous devices for determining glucose have been proposed, but most of these are limited in terms of reproducibility and response rate, as well as available glucose concentrations. Some of the commercially successful commercial methods are based on the use of peroxide, as described above, wherein glucose is the substrate and the oxidized product is glucono-1,5-lactone. Other methods are based on peroxide secondary reactions (e.g., colorimetric methods) or a physicochemical assay such as conductivity. However, these are generally slow-reacting and have the disadvantage that they are quite sensitive to the oxygen pressure of the sample, which can vary considerably; at low oxygen pressure, the upper limit of the linearity of the current response may be lower than what is desired for simple accurate measurements. The same considerations apply to measurement methods for non-glucose substrates.

Group (b) - Broker-assisted biosensors

In these devices, the enzyme is kept in a reduced (electron-rich) state as a result of its reaction with a substrate, which is an analyte with a concentration

8851 5 must be specified. A prerequisite for providing a practical sensor is the establishment of an electrical connection between the electron source (certain electron-rich active sites in the enzyme) and the electrode itself. But because the active sites tend to be present in the recesses and folds in the macromolecular enzyme structure, access to them is completely or partially blocked, and thus it is somewhat difficult to provide an electrical connection efficient enough to generate a reliable and sensitive signal. However, the transfer of electrons between the enzyme and the electrode can be facilitated by the inclusion of an electron carrier or mediator which, in oxidized form, receives electrons from the enzyme and then, in a reduced state, transfers them to the electrode where it is reoxidized.

The use of mediators can be illustrated by the recently described biosensors using glucose oxidase immobilized on a carbon electrode. One embodiment uses a covalently bound enzyme immobilized by the cyanuric chloride method (Jonsson and Gorton, 1985, Biosensors, 1, 355-369) which is claimed to provide good stability (several months). However, the sensor has serious drawbacks in that the mediator used, N-methylphenazinium ion (phenazine methosulfate), is unstable and is also easily washable and must be replaced daily in use. The electrode is also sensitive to oxygen concentration, although it was shown that electrochemical transmission through the mediator is competitive with the oxygen reduction reaction. Another biosensor, which also has immobilized glucose oxidase, uses ferrocene or a derivative thereof as a mediator: Cass et al., 1984) Analyt. Chem. 56, 667-673 and EP-A-0 078 636. The transfer of electrons to the electrode via an intermediary takes place as follows:

8851 5 glucose + enzyme [oxidized] ---> glucono-1,5-lactone + enzyme [reduced] enzyme [reduced] + ferrocene [oxid.] ---> enzyme [oxid.] + (Ferrosinium ion) ferrocene .] ferrocene [only] - (electron for electrode) -> ferricium ion

The mechanistic details of the operation of this electrode are not clear: in particular, it is not explained how the highly insoluble reduced form of ferrocene transfers charge to the electrode to maintain cyclic mediator activity (this remark may not be possible against ionic ferrocene derivatives). In addition, its response is rather slow given the potentially very rapid response that would be expected based on the known rates of the enzyme reactions in question, and the life of the electrode is limited due to the limited stability of the enzyme.

There are several disadvantages associated with the use of a mediator in signal transduction: the possibility of it dissolving out of the biocatalyst-containing region, the limited diffusion of oxidized and / or reduced forms, and the instability of the mediator itself.

Group (c) - Direct electron transfer (DET) biosensors

The possibility of constructing a biosensor without the inclusion of an intermediary has been proposed in a recent bioelectrocatalysis review: Tarasevich (1985) Bioelectrochemistry 10, 231-295. Such devices may be referred to as reagent-free or mediator-free. Examples of mediator-free enzyme electrodes are mentioned in Tara

8851 5 sevich, but use organic polymers, e.g., containing methyl viologene-like moieties and / or conductive organic salts such as NMP⁺TCNQ- (N-methylphenazinium-tetrazyano-4-quinodimethane), which modify the properties of the electrodes and fulfill the functions of mediators. Many methods in which electron transfer occurs from redox proteins via modified electrodes fall into this group.

The instability inherent in many organic polymers and salts is known. Thus, the activity half-life of the NMP / TCNQ-modified electrode used in the alcohol biosensor is about 15 days. Such electrodes are also oxygen sensitive.

Based on the published data, it appears that so far only a few purely mediator-free enzyme electrodes have been developed, although many failed attempts are known, most often using carbon-based electrodes. Recent literature on the use of glucose oxidase (Jonsson and Gorton, loc. cit.) suggests the biggest problem in the immobilization of a latent enzyme, which appears to inhibit its electron transfer capabilities due to etheric or other limitations, thus making the inclusion of a mediator necessary.

There are some rare examples of highly active oxidases immobilized on carbon or platinum. For example, Ianiello et al (1982) Analyt. Chem. 54. 1098-1101 describe mediator-free sensors in which glucose oxidase and L-amino acid oxidase are covalently bound to a graphite electrode by the cyanuric chloride method. However, enzyme electrodes have a limited lifespan of 20 to 30 days: Ianiello and Yacynych (1981) Analyt. Chem. 52, 2090-2095. No information is given on the oxygen sensitivity of the electrodes.

Numerous biosensors, particularly glucose sensors, operating in accordance with the above principles have been described in the past, and a representative selection of these has already been presented; for the purposes now in question, one publication must be considered particularly relevant, namely: Matsushita Electric Appliance Industry Company, Japanese Unexamined Patent Publication No. 56-163447. Described here is an indirect glucose electrode, i.e. one in which the hydrogen peroxide produced by the oxidation of glucose in the presence of glucose oxidase: glucose + O 2 ---> gluconolactone + H2O2 enzyme is oxidized on the surface of the platinum electrode:

Η2θ2 ------> ^{2h +} + ²e^- + 02 to produce an oxidation current proportional to the concentration of substrate (glucose) in the sample. The electrode comprises an electrically conductive carbon support which carries a layer of immobilized enzyme, e.g., immobilized glucose oxidase. The electrically conductive support is itself a cast graphite containing up to 10 parts by weight of a fluorocarbon resin as a binder and on which a thin (less than 1 .mu.m) film of platinum has been deposited, e.g. by electrolytically or by steam. The invention is said to avoid the problems associated with immobilizing the enzyme directly on the platinum surface and to provide an enzyme electrode which is claimed to have fast response times (5 seconds), high sensitivity and durability. However, recent experiments with such electrodes have not confirmed these advantages.

Thus, there remains a need for an enzyme electrode, especially, but not exclusively, for use

Ί

8851 5 in glucose biosensors that is reliable and reproducible, has a fast response time and high sensitivity, and has long-term stability.

According to the invention, a new carbon substrate is used for the enzyme electrode, which allows an enzyme, for example glucose oxidase, to be attached to the electrode in a more advantageous manner, and which allows the construction of an amperometric sensor with much better response and stability. This improved enzyme electrode does not require the use of a mediator reagent (although it may be added if desired) and has been found to operate in the presence of very low concentrations of dissolved oxygen. It gives a high response, for example, current densities of hundreds of microamperes per cubic centimeter (apparent electrode area) in a 10 mM glucose solution; this is believed to be much higher than any other prior amperometric enzyme biosensor and can be advantageously used in the manufacture of microsonde biosensors less than 1 mm²electrode area of ​​0 to 100 nanoamperes. The electrode can also be constructed using very small amounts of immobilized enzyme. It responds to glucose much faster than any other known glucose sensor, typically in 1-2 seconds in the absence of a protective membrane, and in 10 to 30 seconds with the membrane. It has considerable stability when stored wet, even at room temperature; the electrodes also have a good response after several months. They have an extended operating range, require a significantly lower operating potential (325 mV compared to the more common 650 mV), and have a remarkably low background in working potential.

The present invention is based on an enzyme electrode or biosensor capable of responding amperometrically to the enzyme.

8851 5 catalytic activity in the presence of its corresponding substrate and comprising an oxidoreductase enzyme immobilized or adsorbed on the surface of an electrically conductive support member which is or contains a porous layer of resin-bonded carbon or graphite particles and which particles are in contact with the elemental metal platinum. The enzyme electrode is characterized in that the elemental platinum group metal is in the form of finely divided particles uniformly dispersed over the thickness of the resin-bonded carbon or graphite layer, and that said particles are pre-adsorbed on the surface of carbon or graphite particles before binding with said resin.

Thus, in clear contrast to the layered, non-heterogeneous platinized carbon supports disclosed in Japanese Laid-Open Application No. 56-163447, the electrode of the invention consists of or contains substantially heterogeneous layers of resin-bonded carbon or graphite particles, said platinum group metal being substantially uniformly dispersed throughout. The resin-bonded carbon powder layer is preferably formed of resin-bonded carbon powder particles to which colloidal platinum or palladium has been precipitated or adsorbed prior to casting to form a substrate. Preferred resin binders for use in casting platinum-plated carbon particles to make the electrode substrate used in the invention are fluorocarbon resins, especially polytetrafluoroethylene.

8851 5

As the carbon powder, any suitable carbon or graphite powder can be used which readily allows subsequent immobilization of the enzyme, and for this purpose carbon powders with a high density of functional groups such as carboxylate, amino sulfur groups on the surface should be used, as opposed to more vitreous bind enzymes only poorly. The particle size can range from 3 to 50 nm, usually from 5 to 30 nm.

Platinum (or palladium) can be applied to the carbon particles in any suitable manner, e.g., by vapor phase precipitation, electrochemical doping, or simply adsorption from a colloidal suspension (which is preferred) to provide a platinum group metal loading of 1 to 20% by weight based on carbon, preferably 5 to 15%. However, these limits are mostly practical, not critical. Less than about 1% of platinum group metal, the output signal drops to a level that is too low to measure for practical purposes except using a very sensitive device. Over about 20%, the load on the platinum group metal becomes uneconomical with little additional benefit in terms of response time, sensitivity, etc. In fact, when the metal load is very high, the sensitivity begins to decrease. In a preferred method, the carbon powder is platinum plated or palladiumized by oxidative decomposition of a platinum or palladium compound such as chloroplatinic acid, or more preferably an oxidizing ligand-containing platinum or palladium complex in the presence of carbon powder to apply colloidal , US-A-4,044,193 and US-A-4,166,143.

After platinumization or palladiumization, the platinumized or palladiumized carbon powder is formed using a suitable water-repellent bonding resin, preferably a fluorocarbon resin such as polytetrafluoroethylene, to produce either a fully self-supporting porous shaped structure, bonded to an electrically conductive substrate, for example metal, carbon or graphite. A particularly preferred substrate material for the shaped, resin-bonded platinum-plated carbon layer is carbon paper, as disclosed in U.S. Pat. No. 4,229,490, or open-cell carbon fabric, as disclosed in U.S. Pat. No. 4,293,396. To maintain maximum porosity, the amount of resin used as a binder should be as small as possible to provide mechanical integrity and stability of the electrode layer, with the thickness of such a layer rarely greater than about 0.1 to 0.5 mm, although higher thicknesses may be used. Taking into account the requirements for structural integrity, mechanical strength and porosity, the amounts of binder resin are not critical and can vary from as little as 5 or 10% by weight, based on the amount of platinum-plated or palladium-on-carbon powder, up to 80% generally in the range of 30 to 70% by weight. A wide variety of resins can be used, including conductive or semiconducting resins, but synthetic fluorocarbon resins, especially polytetrafluoroethylene, are preferred. Given the small but necessary oxygen demand in the oxidation process, it is essential that the binder resin be oxygen permeable. For this purpose, the solubility of the binder in relation to oxygen should be as low as possible at an atmospheric pressure of at least 2 x 10^_3 cm³ 02 (measured at standard temperature and pressure) cm of polymer³ towards.

8851 5

Suitable binders and their known oxygen solubilities according to The Polymer Handbook (Ed. J. Brandrup and EH Immergut) 1 edition (1967), Interscience, are:

Polytetrafluoroethylene (PTFE) Fluorocarbon polymers other than PTFE Polyethyl methacrylate

Polystyrene Polyvinyl acetate Polyvinyl chloride

polycarbonate

Poly (4-methylpentene-1) Polyisoprene

polychloroprene

Poly-1,3-butadiene Silicone Rubber

S x 10² (Cm ^)

0.276 shift. 0.2 up

8,6

18.2 (calculated)

6,3

2,92

0,51

24.3

10.3

7,5

9,7

31,1

More preferred enzyme electrode substrates for use in accordance with the invention are in fact commercially available materials sold under the trademark Prototech, Prototech Company, Newton Highlands, Massachusetts, and have previously been used as electrocatalytic gas diffusion electrodes in fuel cells. The preparation of such materials is described in detail in US-A-4,044,193, US-A-4,166,143, US-A-4,293,396 and US-A-4,478,696, which are incorporated herein by reference in their entirety. Briefly, however, colloidal platinum having a particle size in the range of 15 to 25 Angstroms (1.5 to 2.5 nm) is adsorbed on the surface of the carbon powder (particle size 50 to 300 Angstroms: 5 to 30 nm), for example by forming a platinum sol in situ in the presence of powdered carbon which acts as a sol inoculum. The platinum-plated carbon particles are then cast on an electrically conductive support12

8851 5 structure, e.g. a sheet of carbon paper, using a synthetic resin binder, preferably a fluorinated hydrocarbon resin, and in particular polytetrafluoroethylene.

According to an alternative described in US-A-4,293,396, the platinum-plated carbon particles are impregnated into a preformed porous carbon fabric and bonded thereto using a fluorocarbon resin, preferably polytetrafluoroethylene. It is to be understood, however, that the present invention is not limited to the use of Prototech materials, but encompasses other substrate materials of the same type containing resin-bonded and shaped platinum-plated or palladium-on-carbon powder. In particular, it is contemplated that materials described as fuel cell electrodes in U.S. Pat. No. 4,229,490 may also be used, i.e., carbon paper electrodes having a carbon paper support member, preferably impregnated with a water-repellent resin such as a uniform layer of platinum black and carbon or graphite particles, which particles are bound by a water-repellent binder, in particular polytetrafluoroethylene.

Immobilization of the enzyme on the surface of a resin-bonded, platinum-plated or palladium-on carbon substrate can be performed using various known immobilization methods, for example, covalent attachment with a carbodiimide or carbonyldiimidazole reagent, covalent attachment with 1,6-dinitro-3,4-difluoro .

Typical exemplary guidelines for immobilizing an enzyme, glucose oxidase, are as follows:

8851 5

A. Carbodiimide treatment:

1. Cut appropriately sized pieces of electrode from a sheet of Prototech electrode material.

2. Immerse the electrodes in ethanol for approximately 5 minutes to ensure complete wetting of the PTFE-coated binder and body.

3. Remove the electrodes from the ethanol and wash thoroughly with distilled water to remove residual ethanol.

4. Prepare 5 ml (or less) of a 0.15 M solution of 1-cyclohexyl-32-morpholino) carbodiimide p-methyltoluenesulfonate in 0.1 M pH 4.5 acetate buffer and place the electrodes here for 90 minutes at room temperature. Gentle mixing with a mechanical stirrer can be used. If the electrodes float on the surface of the solution, they are not sufficiently moistened and the treatment should be repeated from step 2.

5. Remove the electrodes and wash them thoroughly with distilled water. Place them in a freshly prepared glucose oxidase solution (5.0 mg / ml) in pH 5.6 acetate buffer for 90 minutes at room temperature with gentle mechanical stirring.

6. Remove the electrodes from the enzyme solution and rinse them thoroughly with 0.1 M acetate buffer. The electrodes are now ready for use.

7. Store the electrodes at 4 ° C in 0.1 M pH 5.6 acetate buffer.

B. Carbonyldiimidazole treatment

1. Perform step 1 above and omit steps 2 and 3.

2. Prepare a solution of N, N'-carbonyldiimidazole in anhydrous dimethylformamide (40 mg / ml).

3. Place the electrodes in this solution for 90 minutes at room temperature while gently stirring mechanically.

4. Remove the electrodes from the solution and dry off the excess carbonyldiimidazole solution before placing them in the freshly prepared glucose oxidase solution for an additional 90 minutes.

5. Perform steps 6 and 7 above.

C. DFDNB treatment

1. Perform steps 1-3 from A above.

2. Wash the electrodes thoroughly in sodium borate buffer (0.1 M pH 8.5).

3. Prepare a solution of 1,6-dinitro-3,4-difluorobenzene in methanol (0.1021 g / 5 ml) and place the electrodes here for 10 minutes at room temperature.

4. Remove the electrodes and wash them thoroughly in borate buffer before placing them in the glucose oxidase solution for an additional 90 minutes at room temperature.

5. Perform steps 6 and 7 from A above.

Other types of coupling agents can be used in the immobilization process, including bifunctional agents of varying chain length, for example, diimidates such as dimethylmalonimidate or dimethylsuberimidate.

Alternatively, it has been found that simple adsorption of an enzyme on a resin-bound platinum-plated or palladium-on carbon powder support, i.e. without cross-linking, is effective with some enzymes, and in particular glucose oxidase.

Usually, but not necessarily, the surface layer of the immobilized enzyme is physically protected by a suitable porous film, for example a polycarbonate film or

8851 5

membrane, which must, of course, be permeable to the enzyme substrate (glucose) to be determined. Such membranes have some disadvantageous effects in increasing the response time of the sensor, but nevertheless even with such a membrane, these sensors have response times comparable to conventional enzyme electrodes and in some cases considerably better.

As already mentioned, the invention relates in particular to glucose oxidase electrodes, i.e. those in which the immobilized enzyme is glucose oxidase, but it is clear that other oxidoreductases can be used, although not always with equivalent efficiency. This is not necessarily due to the inefficiency of the enzyme itself, but to other factors. For example, in the determination of oxalic acid using oxalate oxidase, the oxalic acid substrate itself undergoes electrochemical oxidation at the parent electrode, thus largely masking the potential effect of the enzyme. Other suitable oxidoreductases include lactate oxidase, galactose oxidase, cholesterol oxidase, and other peroxide-producing enzymes, as well as combinations of immobilized enzymes, including combinations of non-oxidizable and oxidase-active to produce proportional to the concentration of that substrate. One such combination is a combination of beta-galactosidase and glucose oxidase (for the quantitative determination of lactose), or a combination of a beta-glucan depolymerizing enzyme, beta-glucosidase and glucose oxidase (for the determination of beta-glucans).

8851 5

Other types of sensor uses involve the use of enzymatic or non-enzymatic reagents or methods that react with the primary substrate of interest in a precursor reaction, wherein the resulting product contains a substance which in turn acts as a substrate for the enzyme electrode of the invention. Several examples of such precursor steps can be found in the field of immunochemical reactions, and methods for using such reactions in the construction of sensors, including immunosensors, using the enzyme electrodes of the invention are known to those skilled in the art.

However, the preferred use of the electrodes according to the invention as biosensors for the detection and / or quantification of oxidizable substrates, in particular glucose, in a sample, in particular a clinical sample such as blood, serum, plasma, urine, sweat, tear fluid and saliva.

Other potential non-clinical applications include:

(a) monitoring of the fermentation process (b) industrial process control (c) environmental monitoring, eg control of waste and pollution of liquids or gases (d) food testing (e) veterinary applications, in particular applications related to the clinical applications proposed above.

If bios and other sensors containing the enzyme electrode material of the invention may contain other structural elements, electrical wires, electrically non-conductive (insulating) supports or probes, etc., such elements in the structure are conventional and not

8851 5 they need to be explained in detail. If, as is usually the case, the electrode material is a paper-thin sheet or strip, it is sufficient that the biosensor generally includes an insulating support member or probe that supports the electrode material and allows the electrode material to be placed in the sample. In such cases, the actual size of the piece of electrode material may be very small, no larger than a few square millimeters, or even smaller. Electrical contact with the electrode material can be effected in a number of ways, for example by mounting the electrode material surface-to-surface in contact with an electrically conductive contact or terminal, e.g., platinum, silver or some other suitable conductive material. In cases where the electrode material is thick and strong enough to be fully self-supporting, insulating support members or carriers for the electrode material may be omitted and the electrical wires connected directly to the surface of the electrode material.

Non-carbon paper support members can be used as such an electrically semiconducting surface, for example a channel transistor (FET) surface, or an electrically non-conductive surface. In the latter case, electrical contact can be made directly to the resin-bonded carbon or graphite layer of the platinum group metal.

The preparation of the enzyme materials of this invention and their properties are illustrated in the following examples.

Example 1: (Comparative) (Prior art)

A prior art enzyme electrode was prepared by electrolytically doping a thin layer (<1 μm) of platinum on the surface of an electrically conductive support comprising

8 51 5 of porous resin-bonded carbon paper having conductive carbon black granules (Vulcan XC-72) having a nominal particle size of 30 nm and cast on a commercially available graphitized carbon paper sheet using 10% by weight of polytetrafluoroethylene as a binder.

Glucose oxidase from Aspergillus niger was immobilized on the surface of various samples of platinum-plated carbon paper using the above-mentioned carbodiimide treatment, and cross-linked with glutaraldehyde by treating the platinum-plated surface of the electrode with an aqueous glucose oxidase solution, drying and then cross-linking.

For subsequent testing, the electrode material was cut into 2 mm diameter discs.

Example 2; Glukoosielektrodi

Glucose oxidase from Aspergillus niger was immobilized on platinum-on carbon paper sold under the tradename Prototech, Prototech Co. Massachussetts, USA, comprising platinum-plated carbon powder particles (Vulcan XC-72) prepared according to Example 1 of US-A-4,044, 193 by doping colloidal platinum (particle size 1.5 to 2.5 nm) with carbon powder (nominal particle size 30 nm) oxidative decomposition of complex platinum sulphite acid (II) using H2O2, followed by casting and bonding of platinum-on-carbon powder to commercial, on the surface of graphitized carbon paper using about 50% by weight of polytetrafluoroethylene. The amount of platinum in the final product is 0.24 mg cm^-2.

Glucose oxidase was immobilized on various samples of Prototech material by the above treatment methods

8851 5, i.e. by treatment with carbodiimide, carbonyldiimidazole and DFDNB treatment.

In separate experiments, glucose oxidase was immobilized on Prototech material by cross-linking glutaraldehyde, and by simple adsorption, i.e. without cross-linking, suspending Prototech material in a freshly prepared glucose oxidase solution (5.0 mg ml-1) in pH 5.6 acetate buffer for 90 minutes at room temperature. Alternatively, the adsorption of the enzyme can be suitably achieved by an electrophoretic method, for which purpose the electrode carrier material is suspended with a positive potential in the enzyme solution for 60 minutes.

Example 3;

Using the carbodiimide treatment described above, the following enzymes were immobilized on platinum-on carbon paper from Prototech, i.e., PTFE-bonded carbon paper made from pre-platinum-on carbon powder (US-A-4,044,193):

lactate oxidase galactose oxidase glucose ioxide i / beta-Galactos idaase i.

To further illustrate the advantages of the invention and the properties of the enzyme materials of the invention compared to previously known electrodes, enzyme materials prepared according to the above examples were tested for amperometric response in a cell with a modified Rank oxygen electrode system (Rank Brothers in Bottisham, Camhe Acta, 183 (1986) 59-66. In this system

8851 The membrane is replaced by a carbon paper enzyme electrode (5 mm dia.) According to the invention attached to a platinum button electrode. A counter electrode (platinum film) was installed through the cell cover. A silver-silver chloride electrode was used as a control. Some tests (using a protective membrane, and an immiscible solution) used a 2-electrode structure; the counter / reference electrode was a surrounding chloride-doped silver ring. In general, the test solutions in pH 7.0 buffer were magnetically agitated while the working electrode was kept at a potential of 600 mV relative to the comparison by means of a potentiostat. When using a 2-electrode structure, a potential of 325 mV was used. The background current was first allowed to fall to a low level for a sufficient time, after which the substrate solution was injected. The current response was recorded on a recorder.

The results obtained are described in detail below and are shown graphically in the accompanying drawing, in which Figure 1 shows the response of a glucose oxidase electrode according to the invention compared to glucose oxidase immobilized on other types of carbon electrodes;

Fig. 2 is a first diagram showing the stability of a glucose oxidase electrode;

Fig. 3 is another schematic diagram showing the response of a glucose oxidase electrode to different glucose concentrations;

Fig. 4 is a diagram showing the response of a glucose oxidase electrode at varying ambient oxygen pressures;

Fig. 5 is a diagram showing the effect of storage at room temperature on the glucose oxidase electrode;

8851 Fig. 6 is a comparative diagram illustrating the effect of storage at room temperature on a prior art electrode;

Figure 7 shows a comparison between the responses of a glutaraldehyde immobilized glucose oxidase electrode of the present invention and a prior art glutaraldehyde immobilized glucose oxidase electrode;

Figure 8 corresponds to Figure 7 but uses carbodiimide immobilization;

Figure 9 shows a comparison between the responses of a carbodiimide immobilized lactate oxidase electrode according to the invention and a carbodiimide immobilized lactate oxidase electrode according to the prior art;

Figures 10 and 11 show the response profiles of the galactose oxidase and lactate oxidase electrodes of the present invention, respectively;

Figure 12 shows the response profile of a combined glucose oxidase / beta-galactosidase electrode according to the invention;

Figure 13 shows the response profile of a glucose oxidase electrode according to the invention using polyvinyl acetate as a binder for platinum-plated carbon powder instead of polytetrafluoroethylene;

Fig. 14 shows a response profile of a glucose oxidase electrode according to the invention, in which glucose oxidase is immobilized on a carbon paper electrode comprising a surface layer of a resin-bonded (polytetrafluoroethylene) palladium-on-carbon powder;

Fig. 15 shows a modified Rank electrochemical cell used to determine the operating characteristics of the electrodes of the invention;

Figure 16 shows a two-electrode structure used for some assays.

Referring first to Figure 15, much of the information presented herein was obtained using the electrochemical cell of Figure 15. This comprises a two-part cell with a base (1) and an annular jacket (2) defining a water chamber (h) through which water can be circulated to control the temperature of the cell, the two parts being connected by an adhesive threaded collar part (3). Central to the base (1) is a platinum contact (d) on which a test disc (a) of paper electrode material containing immobilized enzyme is mounted and held in place on the platinum contact by O-ring rubber seals (e) and (f) when both parts of the cell are connected.

At the top of the cell, which of course will contain the enzyme substrate solution, is placed a plug (4) supported by an adjustable collar portion (g) and fitted with a platinum counter electrode (b) and an Ag / AgCl reference electrode (c). As mentioned, the tests were performed by holding the working electrode at 600 mV, whereby the current was measured from an electrode with an apparent substrate-exposed area of ​​0.14 cm²· The results are shown in the figures as current density, i.e., current power per surface area of ​​the electrode exposed to the substrate.

Referring to Fig. 16, the platinum contact (B) is surrounded by the reference / counter electrode (C) and separated therefrom by an insulating sleeve (G). A porous polycarbonate membrane mounted on the O-ring is used to hold the test disc (E) (paper electrode material containing immobilized enzyme) by a platinum contact. The open sample chamber (F) allows samples to be dropped onto the membrane. The electrode cell is polarized at 325 mV and the current is monitored with a potentiostat (A). 325 The use of a 2-electrode structure at mV has advantages over a standard 3-electrode cell at 600 mV,

8851 5 measures it is easier to use and has a lower background current. However, the choice of one system over another does not affect the performance characteristics of the electrodes of the invention, such as storage stability, operational stability, response linearity, or oxygen dependence.

The results obtained are explained in more detail below.

Linearity and time dependence of responses

Figure 1 of the accompanying drawing shows typical examples of the electrode response with gradual addition of glucose to final concentrations in the range of 0-35 mM, using a 3-electrode cell and with agitation. All three electrodes A, B and C had glucose oxidase immobilized by the above-mentioned method A. Electrode A comprised a cast sheet of an activated platinum support of the invention, i.e., a resin-bonded (polytetrafluoroethylene) platinum-on carbon powder sold under the trademark Prototech; electrode B comprised an electrically conductive support cut from a portion of a graphite rod; electrode C comprised an electrically conductive support cut from commercially available eiplatinized carbon paper. As can be seen, electrodes B and C gave smaller, relatively slow responses, reminiscent of the results obtained with mediator sensors commonly reported in the literature. Electrode A gave more reliable and stable responses with a response time of about 1 second. (The peak in the signal observed at the top of the initial response is insignificant and is partly a consequence of the injection method; the glucose-dependent plateau is the signal of interest). All three electrodes gave a relatively linear

8851 5 responses to glucose concentration (Figure 2, results shown for A, C only). This covers the area required for direct analysis of glucose in the blood (1 30 mM). Similar results obtained with type A electrodes using immobilization method B suggest that this method gives even better linearity over an extended range.

As seen in Figure 2, the response of electrode A was virtually unchanged after 23 days, but the responses of the others deteriorated over time (as shown for electrode C). This type of behavior was also observed for the other immobilization methods described above, all of which could be used to make sensitive and stable electrodes with the carbon material used for A, but gave unsatisfactory electrodes with many other inactive carbon materials. The response time of A was also unchanged after 23 days, while the other electrodes showed an increase in the response time, with initial response times of about 23 to 30 seconds, which increased after 8 days to 2-3 minutes. Active electrodes (such as A) generally showed some decrease in response during the first day, but the response then reached a level relative to time. Type A electrodes stored wet (pH 5.6) at 4 ° C and tested periodically over a 6 month period changed little (after the first days) and although there was a slight gradual deterioration after this time, the response after 12 months was still 70% of the original value.

Figures 1 and 2 show the current of the electrode / u cm “2 at an operating potential of 600 mV using the

3-electrode.

8 51 5

The extended shelf life and stability of the electrodes in question are illustrated in Figure 5, which shows the response of a carbodiimide-immobilized glucose oxidase electrode to 5 mM glucose after storage in pH 5.6 acetate buffer at room temperature for a period of 180 days. Figure 6 shows the comparative results for a prior art electrode (Example 1). Measurements in Figure 5 were made at 600 mV with a 3-electrode system and at Figure 6 325 mV with a 2-electrode structure.

Further comparisons between the prior art electrodes (Example 1) and the electrodes of the invention using different enzymes and different immobilization methods are shown in Figures 7-9, where all measurements were made at 325 mV.

The response curves for lactate, galactose and lactose (combined glucose oxidase and beta-galactosidase) are shown in Figures 10 to 12. These measurements are made at 600 mV.

Electrode Reuse s Suitability for repeatable assays

Under continuous loading, the electrodes of this invention exhibited exceptional durations not previously shown. This was illustrated by the following series of demanding experiments.

First, a glucose oxidase electrode (Example 2) was placed in a closed cell and allowed to react amperometrically with the agitated glucose solution (initial concentration 5 mM, initial current 100 / μm). It was used continuously for 18 hours, during which time the signal gradually decreased

8851 5 ki to less than 10 / uamp. The total electricity generated was about 75% of theoretically expected (based on 2 donated electrons per molecule of glucose). This experiment was immediately repeated with the same electrode by regenerating the glucose solution, after which the initial current was set again and the determination of the substrate under continuous loading gave identical results.

In the next experiment, power generation using the same electrode was continued for another 5.7 days, but by feeding the glucose solution by circulating from a large tank to maintain a concentration of 5 mM. The power gradually decreased over 100 hours but then stabilized at 45 / uA, which was maintained for another 40 hours. It is possible that small amounts of loosely bound enzyme were released from the electrode base during this time (although there was no dependence of current power on the stirring rate), or that some other influencing factor is involved.

After the long-term tests described above, the electrode response was tested at different glucose concentrations as mentioned above. Although the amplitude of the signal was lower than that of the freshly made one, the electrode gave a very satisfactory step function in the range of 0 to 30 mM glucose, which confirms that it has not suffered any detrimental effects due to continued use under load. This conclusion was confirmed in three additional tests after 1 week of storage (4 ° C), 8 weeks of storage, and again after use under load for an additional 4.7 days: responses to different glucose concentrations were unchanged.

These tests show that the electrode could be used for a total of at least 250 hours (over 15,000 minutes), thus giving the electrode material a potentially very high

8851 5 long service life. The lifespan of prior art enzyme electrodes is generally much shorter than that of the invention, in some cases only a few hours: Turner, (1985) Proceedings Biotech 85 (Europe) Online Publications, Pinner UK 181-192. For example, glucose electrodes based on ferrocene-coupled glucose oxidase generally have half-lives of only about 24 hours (Turner, loc. Cit.), Whereas Cass et al. (1984) Analyt. Chem. 56, 667-673 show a permanent lifetime of a total of about 50 hours for the same electrodes. When the electrode was used for 50 consecutive measurements of 5 mM glucose solution, the standard deviation was less than one percent.

Suitability for continuous monitoring

The final response level measured in the above-mentioned reuse experiments (45 μg at 5 mM glucose) remained unchanged upon exposure to aerated solutions and was the same after several weeks of additional storage. The current stability of this electrode over 12 hours was checked in a control test using a sterilized glucose solution under conditions designed to eliminate possible glucose losses due to bacterial contamination. The signal was constant throughout the period, suggesting that the potential effects of electrode initial conditioning in the agitated and recycled solution over several days were complete. Such electrodes, suitably conditioned by this or any other suitable conditioning / washing method, have use where continuous glucose monitoring is required.

8851 5

Batch repeatability

Provided that suitably clean manufacturing conditions are maintained, all electrodes made in accordance with this invention will function as shown, providing a good response to glucose. Pairs of electrodes of identical size and similarly prepared gave very consistent results (current responses within a few percent when tested under identical conditions). In addition, all the electrodes thus prepared had very long durations and durations, as mentioned above, compared to the prior art electrodes. Thus, the electrodes of the invention can be stored reliably and used for several weeks, while the electrodes of the prior art often varied in the same batch. For example, Turner, loc. cit., found that although a few glucose electrodes in one batch were exceptionally long-lived (half-life 600 hours), most had half-lives of about 24 hours. Thus, they could not be used reliably for much more than 24 hours.

Dependence of response on oxygen concentration

To test the effect of dissolved oxygen, a test cell was modified to include an oxygen electrode in addition to the glucose electrode. Experiments were performed in which dissolved oxygen was purged out of the system with argon. Under these conditions, the type A (above) electrode gave a rapid response to glucose additions, suggesting a mechanism that is largely independent of the surrounding oxygen concentration. Such a result due to the special surface structure properties of the electrode combined with favorable enzyme immobilization has not been observed before.

8851 5

In another experiment (Fig. 4), the output signal of a type A electrode (prepared by Method A above) at 600 mV was continuously monitored by purging with argon. At the same time, the oxygen pressure in the sample was measured. The results shown in Figure 4 show a substantially constant current signal (upper curve) that is substantially independent of the oxygen pressure of the sample solution (lower curve). In the second experiment, the signal was virtually independent over a period of 10 minutes, while the oxygen pressure decreased rapidly. With the second electrode (prepared by Method B), the current decreased by less than 5% over a period of 3 minutes during which oxygen was consumed by 90%. An increase in current response is also observed when oxygen is added back to the system, although this effect is relatively slow. When further purging with argon, the electrodes respond only to a limited range of glucose concentrations, and it is possible that the presence of residual oxygen is necessary to initiate the enzymatic activity responsible for capturing hydrogen from the substrate. Nor can it be ruled out that oxygen adsorbed on the electrode (at a low concentration, and not expressed by the oxygen electrode) plays a role in this behavior.

The enzyme loading

Independent measurements of glucose consumption rate and maximum current densities show that the amount of enzyme actively immobilized per electrode (type A) corresponded to about 7 μg of active enzyme per cubic centimeter of electrode surface. (There is little data in the literature on the enzyme loads of similar glucose oxidase-based biosensors). In the immobilization methods, it was found that even after diluting the enzyme solution 10-fold, highly active electrodes could still be prepared.

8851 5

Temperature dependence of the response

Type A electrodes were tested in the concentration range of 0 to 30 mM glucose at temperatures of 10 to 37 ° C. The temperature coefficient was 2 to 3% per degree (corresponding to an Arrhenius activation energy of about 24 kJmol^-1). This can be compared to 4% per ° C given to a ferrocene-mediated biosensor (Cass et al., Loc. Cit.).

The pH dependence of

A slight pH dependence of the response was observed. However, between pH 7.0 and 8.0, the response was virtually independent of pH, except at very high glucose concentrations (025 mM).

Response of the electrode covered by the protective membrane

It was found that the polycarbonate membrane caused little change in the form and intensity of the electrode response in the agitated system. The response time in the undisturbed system was about 20 seconds.

Use of an electrode to analyze whole blood samples

An electrode equipped with a polycarbonate membrane was satisfactorily used for direct measurement of blood glucose. The interference due to ascorbate, 0.2 mmol / liter, was about 2.5% of the total signal at the glucose level of 5 mmol / liter.

8851 5

Electrode use in different structures of analytical biosensors

The successful use of the electrode according to the invention in a Rank-type cell using a modified Clark electrode is confirmed by the above results. It has also been shown that the electrode gives excellent results when used in other types of sensors, such as a probe.

For example, a 2 mm diameter probe of the type commonly used in conventional electrodes was constructed with the electrode mounted on a wire and sealed in a glass tube. This could be immersed (along with connected reference and counter electrodes) in agitated solutions in a dish or other vessel to reliably measure glucose concentration without the need to eliminate atmospheric oxygen. Measurements made with this and smaller probes of similar structure showed that the current response of the electrode in a solution with a certain glucose concentration is approximately proportional to the apparent surface area or weight of the electrode.

Probes were also constructed in which the electrode was minimally small (about 0.25 to 0.50 mm² area, weight 30 60 / ug). The wire assembly was covered with a plastic sleeve and the covered probe was placed in a catheter needle (1.5 mm diameter). The needle can be inserted through a rubber seal into a container (which may be included in a fermentor or similar device, or a waste container) and used as a probe sensor to determine the glucose concentration of the solution in the container. In this construction, the surrounding needle protects the sensor electrode when it is inserted, but can also be pushed out of the needle after insertion if necessary.

Although the aforementioned miniature probes typically gave signals in the range of 1 to 10 μamp, accurate measurements in the range of 1 to 100 namp can be obtained with suitable instrumentation. Since the signal currents in this range are provided by enzyme electrodes (according to the invention) with a very small size (about 0.005 mm²area, weight 1 / μg weight), such electrodes can be placed in fine needle microscopes that can be used in catheter probes for in vivo measurements.

Although not fully understood by the mechanism underlying the operation of the electrodes, certain conclusions can be drawn from the results obtained. Thus, it is known that the presence of active surface groups in carbon formed by surface oxidation at high temperature allows the crosslinking reactions necessary for immobilization of enzymes, and the number and variability of such surface groups are likely to increase when platinum (or other platinum group metal such as palladium) is present. Stonehart (1977), Modern Aspects of Electrochemistry, No. 12, Ed. Bockris and Conway, Plenum Press, New York, 183-266). It is apparent that there are differences in enzyme binding when using different immobilization methods. (For example, many of the published methods use different amino acid residues for enzyme attachment, while cyanuric chloride-activated enzyme-bound enzymes attach only from their lysine residues: cf. Ianiello and Yacynych (1981), Analyt. Chem. 53, 2090-2095). It is not expected that the variations in the tertiary structure of enzymes resulting from immobilization will be identical for all immobilization methods, which may be the reason for the large variations in enzyme activity and stability observed in this type of work.

The highly heterogeneous nature of the base electrode material used in this invention, in contrast to the layered, non-heterogeneous structure of the electrode type described in, for example, Japanese Published Patent Application 56-163447, maximizes the possibility of obtaining numerous different types and orientations in an integrated three-dimensional structure. In the absence of crosslinking reagents, it also allows for strong surface adsorption. The pores in the bound carbon matrix allow the enzyme molecules to enter to surround the components of the matrix, which provide a very large surface area for the enzyme, and allow conformations that are advantageous for its stability and activity. (This is in contrast to bonding to relatively flat surfaces, such as platinum, vitreous carbon, or graphite, which have a much smaller surface area, which places constraints on conformation, as reported in the previous literature). In addition, the very fast response times (1-2 seconds) of the electrodes of the invention indicate a very fast electron transfer to the electrode, which requires not only very high enzyme activity, but is facilitated by a sufficient number of electron receptor sites on the electrode itself. These are achieved by the high density of fine platinum carbon granules distributed over a very large surface area within the microstructure, which provides the highest possible probability that surface platinum growth sites are accessible to the active sites of the enzymes.

To demonstrate the utility of other resins as binders in the enzyme electrodes of the invention, as well as to demonstrate the utility of other platinum group metals, glucose oxidase electrodes have been constructed using polyvinyl acetate as a binder, and palladium as a platinum group metal.

8851 5

In the former case, a glucose oxidase electrode was constructed by immobilizing glucose oxidase on the surface of a platinum-on carbon paper electrode described above by Method A described above (Example 2) but using 50% by weight of polyvinyl acetate as a binder instead of polytetrafluoroethylene.

When tested using the same modified Rank electrode system at 325 mV, a substantially linear response was obtained, as shown in Figure 13.

In the latter case, a glucose oxidase electrode was constructed by immobilizing glucose oxidase on the surface of a palladium-on-carbon paper electrode prepared by the method described above prepared by doping palladium on carbon powder (apparent particle size 30 nm: Vulcan XC-72) and then bonding the palladium-on-carbon powder in a thin layer (0.1 mm) to the surface of the electrically conductive carbon paper using 50% by weight, based on the weight of the palladium-on-carbon powder, of polytetrafluoroethylene as a binder.

A mm disk cut from treated palladium-on carbon paper was mounted on the platinum contact of the 2-electrode cell shown in Fig. 16 and tested for response at 325 mV. The results are shown in Figure 14 and again show a substantially linear response in terms of current density relative to glucose concentration.

Given the clear equivalence of Pt, Pd, Ru, and Rh with other platinum group metals in gas diffusion electrodes according to US-A-4,293,396 and other publications, it is expected that other platinum group metals, e.g., ruthenium and rhodium, will be effective.

8851 5 alternatives to platinum and palladium in the enzyme electrodes according to the invention.